EU Referendum Predictions

Britain’s decision day is here, and today the electorate will decide whether or not the UK should remain within or leave the European Union.

Who’s going to win is anyone’s guess, with the polls being almost dead level going into today’s historic vote, but with the help of some precedent and some stats – I’m going to make a prediction on how things are going to turn out tonight.

Here are how the final polls look going into the vote:

	Remain	Leave	Undecided
Online	43.6%	45.1%	11.3%
Telephone	46.8%	46.6%	6.6%
TOTAL	44.6%	45.6%	9.9%

As you can see, Leave are actually slightly ahead – but what’s crucial is that there are a large number of undecided voters who haven’t made up their mind yet as to which way they’ll cast their vote, and so far in the last few days we’ve seen a clear shift towards Remain in the polls.

It’s these people who will decide which side wins the referendum.

With the current support of both sides, the Remain camp need to win 55% of undecided voters while Leave only need to win 45%.  However, this is much easier said than done.

Undecided voters tend to choose the status quo in elections/referendums, so that’s why my final predictions are going to see the Leave side pick up fewer undecided voters than the Remain side.

In the 2014 Scottish independence referendum, my final polls suggested a 47%-43% lead for the No campaign, with the remaining 10% undecided.  In the results, the “Remain” side in that referendum picked up 8/10 voters and won by a 10% margin.

Similarly, in the 1995 Quebec independence referendum, a polling average of the final polls gave the Yes side a 46%-42% lead but with just over 7/10 voters going to the No camp – they managed to pull off a win by just over 1% of the total vote.

History, then, is not on the side of the Leave camp but with the slightly different nature of this vote and the unusual combination of the older voters being more in favour of Leave than younger voters and older voters being more likely to turnout, then the swing does have a chance of happening, although the chances of it being as strong as they need are unlikely.

I’ve create a polling model based upon a 5/2 split towards Remain, which is ever so slightly more than the Quebec vote turned into but one that I believe will be accurate.  It takes an average of the most recent phone and telephone polls from British Polling Council members, and in the case of the individual nations it weights by the polling organisations that provided samples for them.

So without further ado, here’s how I see the referendum results going:

	Remain	Leave
England	51.2%	48.8%
Scotland	60.7%	39.3%
Wales	54.7%	45.5%
NI	58.4%	41.6%
UK	51.6%	48.4%

 

Remain has the slightly better chance of winning, but with such a close-run vote it could be late in the morning before we finally know the destiny of Britain & Europe’s long and complex relationship.

I very much hope that I’m right and Remain win the day, but it’s now in the British public’s hands and no matter what the result as dawn breaks tomorrow, Britain has turned into a very different place.

Polls consulted

Online

Date	Poll Comp.	Sample	Remain	Leave	Undecided
16-22 Jun	TNS	2320	41%	43%	16%
20-22 Jun	Opinium	3011	44%	45%	12%
20-22 Jun	Yougov	3766	45%	45%	10%
10-13 Jun	ICM	2001	44%	49%	7%

Telephone

Date	Poll Comp.	Sample	Remain	Leave	Undecided
17-22 June	ComRES	1032	48%	42%	11%
20-Jun	Survation	1004	45%	44%	11%
16-19 Jun	ORB	800	53%	46%	2%
11-14 Jun	Ipsos Mori	1257	43%	49%	3%
10-13 Jun	ICM	1000	45%	50%	5%